2. Subtitle
www.Desire2Learn.com
Analytics Primer
What is this thing called Analytics?
– Using statistical analysis to discover or derive meaningful insight
from the data we generate
• 2.5 quintillion bytes of data were produced every day in 2012
• 90% of the data in the world today has been created in the last two years alone
– Data generated from many different sources
– Data can be about any subject matter
3. Subtitle
www.Desire2Learn.com
Analytics Primer
What is Predictive Analytics then?
– Using statistical analysis, machine learning, data modeling and data
mining to discover or derive meaningful insight from the data we
generate
• Historical and current data
• Make predictions about future behaviors or otherwise unknown events
• Exploit patterns found in big data to identify risks and opportunities in a given
subject matter
4. Subtitle
www.Desire2Learn.com
Analytics Primer
So what does this mean for Education?
– Using statistical analysis, machine learning, data modeling and data
mining to discover or derive meaningful insight from the data
students, instructors and institutions generate
• Historical and current data
– Data generated from students, courses, programs, instructors, etc.
• Make predictions about future behaviors or otherwise unknown events
– Who is at risk to drop out? Who is best to recruit? What is the best delivery methodology for a
course?
• Exploit patterns found in big data to identify risks and opportunities in a given
subject matter
– Which students will not graduate? Which students take too long to graduate?
5. Subtitle
www.Desire2Learn.com
Analytics Primer
What kinds of data are in Education?
– Learning and academic data within and across
institutions
• Engagement
– Enrollments, Withdrawals, Logins, Tools Usage, Course Access, Content
Access, Social Learning
• Learning Outcomes
– Competencies, Learning Objectives, Course Outcomes, Program Outcomes
• Assessments
– Grades, Quiz Analysis, Surveys, Rubrics, Risk Analysis
6. Subtitle
www.Desire2Learn.com
Predictive Analytics
What areas of Education will benefit?
– Institutional effectiveness
• Ability to respond to funding reform changes
• Enable new approach to recruitment and retention strategies
• Proactive course content and delivery modifications
– Student Success
• Improve first year retention rates
• Tailor or target course and degree pathways for students
• In-flight course analytics to improve success
8. Subtitle
www.Desire2Learn.com
Educate. Motivate. Graduate.
Real-time, personalized academic progression
and monitoring tool
– Course is active or in-progress
– Student is active within the course
Predictive across key learning domains
– Course and Content Access
– Social Learning/Engagement
– Grades
– Preparedness
– Completion
9. Subtitle
www.Desire2Learn.com
At Risk Does Not Mean Alone
Predictive tool assists Instructors to
– Quickly and easily identify academically
at-risk, disengaged or isolated students
– Understand key learning issues on a
per course and per student basis
– Develop personalized learning strategies
– Target prescriptive resolution
Ongoing advisement and planning tool for both students and
instructors
• Continuous feedback and improvement for adaptive learning
• Minimize any opportunity for failure
10. Subtitle
www.Desire2Learn.comDesire2Learn Internal Use Only
Right Student. Right Course. First Time.
Degree Compass uses predictive analytics to
– Provide grade prediction based on historical success and
student’s capabilities
– Presents ranked list of courses relevant to the student’s degree
pathway
Student retention, graduation, time-to-degree completion
• More informed course choices leads to successful course completion
– Minimizes attrition, credit creep and time-to-degree
– Maximizes degree completion opportunity
Welcome to this session.Focus of session is on HE but application can have roots in K-12 as well. Taking a different tact in this session. Rather than give you a number of industry definitions, we want to use real language and real-world examples of analytics and predictive analytics to show the application in education.
Before we get into the meat of the presentation, I thought I would give a brief analytics primer. Is everyone in this room familiar with what analytics is? Clear as mud? Are you as confused as I have been over what analytics is and what it is supposed to be doing for us or helping us achieve? Boy, do we generate data these days! Impossible to use old school methods of analysis given the sheer quantity of data produced. New analytical methods allow the deeper dive that traditional reviews or assessments are incapable of.Briefly comment on the amount of data we generate in this digital world and the sheer difficulty to fully understand or appreciate what insight can be derived from that data. Examples of subject matter and sources of data:Weather – global and local satellites, ground sensors, etc.Automobiles – geography, make and model, on-board diagnostics, etc.Medicine – age, habitat, diet, genetics, lifestyle, etc.
What we have seen in the previous slides is what the statistical analysis of historical data can do for the mortgage, car insurance and retail industries. But what if we could better predict behavior? While there is some element of predictive in the previous examples, it is at the crowd level. What if we could predict or, better yet, target predictions at the individual level? What if we could predict individual future behaviors or actions from the data we generate? Story of Minnesota man who was going to sue Target for sending his teenage daughter coupons for baby clothes and diapers. Turns out Target is so in tune with their shoppers habits that they knew she was pregnant even before her father did. Clues such as vitamin supplements, large quantities of lotion, and hand sanitizers, typical to many pregnant women according to the Target department, signal other items the consumer may need. The father has since dropped his lawsuit.
Reams of data generated in higher education institutions. From data generated by the learner, to the instructor to the advisor, this data delivers tremendous insight into the learning landscape – the integrity of the learning environment.Each of these data points can provide valuable information and insight into solving problems or providing actionable information to improve the learning experience for the student, instructors and the institution itself. Quality improvement initiative that is tailored to your organization.Quality improved initiative that is tailored to the student.
Reams of data generated in higher education institutions. From data generated by the learner, to the instructor to the advisor, this data delivers tremendous insight into the learning landscape – the integrity of the learning environment. It is all about responding to the indicators in the environment. Predictive analytics gives you the leg up on understanding and responding to the environment before the event occurs (ie. student failing out, student dropping out, course or program losing accreditation because of inability to deliver skilled graduates into the field, etc.)Each of these data points can provide valuable information and insight into solving problems or providing actionable information to improve the learning experience for the student, instructors and the institution itself. Quality improvement initiative that is tailored to your organization.Quality improved initiative that is tailored to the student.
D2L’s Student Success System empowers institutions with predictive analytics tool for improving student success, retention, completion, and graduation rates. It applies predictive analytics to tailor or target the individual student learning environment. By analyzing the student across key learning domains, course specific predictions of the student’s success can be made. Real-time, weekly risk indicators are provided that help identify to the instructor and the student and early signs or patterns of proficiency problems.Student Success System allows a dynamic between the student and instructor that is unprecedented. A holistic view of the student’s current academic progress or academic health while they are in-flight within the course. No longer do they find out at the end or near the end of the course that they are going (or have gone) way off track. Provides help before you even know you may need help. It an academic tutor-in-a-tool. Uses machine intelligence and statistical techniques to identify at-risk students pre-emptively within the first few weeks of the semester. Tool guides and mentors both instructors and students and identifies the key areas that need a resolution in order to achieve success.
Gives students and instructors an unprecedented preview of their success. Assistive tool to help students modify or develop behaviors that lead to success. Helps instructors to understand whether course content and/or delivery methodologies are appropriate. Assists instructors and students with guidance and direction on the progress of they are making while the course is in-flight. It is a mentoring, coaching and consulting tool for both students and instructors. From an institutional perspective, Student Success System Drive helps assess the integrity of the learning environment.Student Success System uses predictive analytics to help students and instructors make more informed choices to lead to successful graduation as well as minimize any opportunity for failure (that exists in the current system).In addition, rich data visualizations help students and instructors very quickly understand where the key learning areas are as well as transform complete student engagement and grade assessments data into easily identifiable patterns of academic risk or weakness – both within and across the class cohort.Student Success System allows the learning momentum to continue unbroken both for the student and the instructor.
DC intervenes at an even earlier stage in the learning pathway. DC provides predictive intelligence and guidance before you even make the course choice.Creates first conditions for success by aligning student’s academic path with their scholastic capabilities. By providing apredicted grade before the course even begins contributes to an overallearly intervention strategy for success.Gives students a simple and easy tool to continue to track, manage and complete their degree on timeGetting students graduated on time reduces decreases the ongoing costs of higher education (wasted tuition dollars) and thereby, student debt Getting students graduated on time reduces loss of income (I have numbers for lost income)Getting students graduated and contributing to society helps improve local and federal economies by providing an educated workforce with increased incomesGetting students graduated and contributing to society helps local and federal budgets reclaim tax dollars spent on education (funding to institutions, Pell grants, student loans (unpaid or otherwise))
Degree Compass and Student Success are the killer one-two punch. They are the first steps in D2L’s predictive analytics product strategyDegree Compass helps you make the right choice on day one. Student Success helps you in-course, through the day-to-day needs DC intervenes at an even earlier stage in the learning pathway. DC provides predictive intelligence and guidance before you even make the course choice.
D2L’s predictive portfolio is at the pleading edge of an emerging industry.Competitors are behind and sitting between Reporting and Forecasting levels at best.